LHS ia a Rails-Gem, providing an ActiveRecord like interface to access HTTP-JSON-Services from Rails Applications. Special features provided by this gem are: Multiple endpoint configuration per resource, active-record-like query-chains, scopes, error handling, relations, request cycle cache, batch processing, including linked resources (hypermedia), data maps (data accessing), nested-resource handling, ActiveModel like backend validation conversion, formbuilder-compatible, three types of pagination support, service configuration per resource, kaminari-support and much more.
LHS uses LHC for advanced http requests.
gem 'lhs'
# config/initializers/lhc.rb
LHC.configure do |config|
config.placeholder(:service, 'https://my.service.dev')
end
# app/models/record.rb
class Record < LHS::Record
endpoint '{+service}/records'
endpoint '{+service}/records/{id}'
end
# app/controllers/application_controller.rb
record = Record.find_by(email: 'somebody@mail.com')
record.review # "Lunch was great
Gemfileconfig/initializers/lhc.rb (See: https://github.com/local-ch/lhc#configuration)LHC::Caching to LHC.config.interceptors to facilitate LHS' Request Cycle Cacheapp/models for autoload/preload reasonsLHS from within your rails controllersEndpoint, the entry point to a service, a process, or a queue or topic destination in service-oriented architecture
Start a record with configuring one or multiple endpoints.
# app/models/record.rb
class Record < LHS::Record
endpoint '{+service}/records'
endpoint '{+service}/records/{id}'
endpoint '{+service}/accociation/{accociation_id}/records'
endpoint '{+service}/accociation/{accociation_id}/records/{id}'
end
You can also add request options to be used with configured endpoints:
# app/models/record.rb
class Record < LHS::Record
endpoint '{+service}/records', auth: { bearer: -> { access_token } }
endpoint '{+service}/records/{id}', auth: { bearer: -> { access_token } }
end
-> Check LHC for more information about request options
It's common practice to use different hosts accross different environments in a service-oriented architecture.
Use LHC placeholders to configure different hosts per environment:
# config/initializers/lhc.rb
LHC.configure do |config|
config.placeholder(:search, ENV['SEARCH'])
end
# app/models/record.rb
class Record < LHS::Record
endpoint '{+search}/api/search.json'
end
DON'T!
Please DO NOT mix host placeholders with and endpoint's resource path, as otherwise LHS will not work properly.
# config/initializers/lhc.rb
LHC.configure do |config|
config.placeholder(:search, 'http://tel.search.ch/api/search.json')
end
# app/models/record.rb
class Record < LHS::Record
endpoint '{+search}'
end
LHS uses endpoint configurations to determine what endpoint to use when data is requested, in a similiar way, routes are identified in Rails to map requests to controllers.
If they are ambiguous, LHS will always use the first one found:
# app/models/record.rb
class Record < LHS::Record
endpoint '{+service}/records'
endpoint '{+service}/bananas'
end
# app/controllers/some_controller.rb
Record.fetch
GET https://service.example.com/records
Be aware that, if you configure ambigious endpoints accross multiple classes, the order of things is not deteministic. Ambigious endpoints accross multiple classes need to be avoided.
You can inherit from previously defined records and also inherit endpoints that way:
# app/models/base.rb
class Base < LHS::Record
endpoint '{+service}/records/{id}'
end
# app/models/record.rb
class Record < Base
end
# app/controllers/some_controller.rb
Record.find(1)
GET https://service.example.com/records/1
In case you want to just fetch the records endpoint, without applying any further queries or want to handle pagination, you can simply call fetch:
# app/controllers/some_controller.rb
records = Record.fetch
GET https://service.example.com/records
You can query a service for records by using where:
# app/controllers/some_controller.rb
Record.where(color: 'blue')
GET https://service.example.com/records?color=blue
If the provided parameter – color: 'blue' in this case – is not part of the endpoint path, it will be added as query parameter.
# app/controllers/some_controller.rb
Record.where(accociation_id: '12345')
GET https://service.example.com/accociation/12345/records
If the provided parameter – accociation_id in this case – is part of the endpoint path, it will be injected into the path.
You can also provide hrefs to fetch multiple records:
# app/controllers/some_controller.rb
Record.where('https://service.example.com/accociation/12345/records')
GET https://service.example.com/accociation/12345/records
In order to reuse/dry where statements organize them in scopes:
# app/models/record.rb
class Record < LHS::Record
endpoint '{+service}/records'
endpoint '{+service}/records/{id}'
scope :blue, -> { where(color: 'blue') }
scope :available, ->(state) { where(available: state) }
end
# app/controllers/some_controller.rb
records = Record.blue.available(true)
GET https://service.example.com/records?color=blue&available=true
You can fetch all remote records by using all. Pagination will be performed automatically (See: Record pagination)
# app/controllers/some_controller.rb
records = Record.all
GET https://service.example.com/records?limit=100
GET https://service.example.com/records?limit=100&offset=100
GET https://service.example.com/records?limit=100&offset=200
# app/controllers/some_controller.rb
records.size # 300
In case your record endpoints are not implementing any pagination, configure it to be paginated: false. Pagination will not be performed automatically in those cases:
# app/models/record.rb
class Record < LHS::Record
configuration paginated: false
end
# app/controllers/some_controller.rb
records = Record.all
GET https://service.example.com/records
The different behavior of count and length is based on ActiveRecord's behavior.
count The total number of items available remotly via the provided endpoint/api, communicated via pagination meta data.
length The number of items already loaded from the endpoint/api and kept in memmory right now. In case of a paginated endpoint this can differ to what count returns, as it depends on how many pages have been loaded already.
find finds a unique record by unique identifier (usually id or href). If no record is found an error is raised.
Record.find(123)
GET https://service.example.com/records/123
Record.find('https://anotherservice.example.com/records/123')
GET https://anotherservice.example.com/records/123
find can also be used to find a single unique record with parameters:
Record.find(another_identifier: 456)
GET https://service.example.com/records?another_identifier=456
You can also fetch multiple records by id in parallel:
Record.find(1, 2, 3)
# In parallel:
GET https://service.example.com/records/1
GET https://service.example.com/records/2
GET https://service.example.com/records/3
find_by finds the first record matching the specified conditions. If no record is found, nil is returned.
find_by! raises LHC::NotFound if nothing was found.
Record.find_by(color: 'blue')
GET https://service.example.com/records?color=blue
first is an alias for finding the first record without parameters. If no record is found, nil is returned.
first! raises LHC::NotFound if nothing was found.
Record.first
GET https://service.example.com/records?limit=1
first can also be used with options:
Record.first(params: { color: :blue })
GET https://service.example.com/records?color=blue&limit=1
last is an alias for finding the last record without parameters. If no record is found, nil is returned.
last! raises LHC::NotFound if nothing was found.
Record.last
last can also be used with options:
Record.last(params: { color: :blue })
After fetching single or multiple records you can navigate the received data with ease:
records = Record.where(color: 'blue')
records.length # 4
records.count # 400
record = records.first
record.type # 'Business'
record[:type] # 'Business'
record['type'] # 'Business'
How to configure endpoints for automatic collection detection?
LHS detects automatically if the responded data is a single business object or a set of business objects (collection).
Conventionally, when the responds contains an items key { items: [] } it's treated as a collection, but also if the responds contains a plain raw array: [{ href: '' }] it's also treated as a collection.
If you need to configure the attribute of the response providing the collection, configure items_key as explained here: Determine collections from the response body
To influence how data is accessed, simply create methods inside your Record to access complex data structures:
# app/models/record.rb
class Record < LHS::Record
endpoint '{+service}/records'
def name
dig(:addresses, :first, :business, :identities, :first, :name)
end
end
Nested records, in nested data, are automatically casted to the correct Record class, when they provide an href and that href matches any defined endpoint of any defined Record:
# app/models/place.rb
class Place < LHS::Record
endpoint '{+service}/places'
endpoint '{+service}/places/{id}'
def name
dig(:addresses, :first, :business, :identities, :first, :name)
end
end
# app/models/favorite.rb
class Favorite < LHS::Record
endpoint '{+service}/favorites'
endpoint '{+service}/favorites/{id}'
end
# app/controllers/some_controller.rb
favorite = Favorite.includes(:place).find(123)
favorite.place.name # local.ch AG
GET https://service.example.com/favorites/123
{... place: { href: 'https://service.example.com/places/456' }}
GET https://service.example.com/places/456
If automatic detection of nested records does not work, make sure your Records are stored in app/models! See: Insallation/Startup checklist
Typically nested data is automatically casted when accessed (See: Access and identify nested records), but sometimes API's don't provide dedicated endpoints to retrieve these records.
In those cases, those records are only available through other records and don't have an href on their own and can't be casted automatically, when accessed.
To be able to implement Record-specific logic for those nested records, you can define relations/associations.
# app/models/location.rb
class Location < LHS::Record
endpoint '{+service}/locations/{id}'
has_many :listings
end
# app/models/listing.rb
class Listing < LHS::Record
def supported?
type == 'SUPPORTED'
end
end
# app/controllers/some_controller.rb
Location.find(1).listings.first.supported? # true
GET https://service.example.com/locations/1
{... listings: [{ type: 'SUPPORTED' }] }
class_name: Specify the class name of the relation. Use it only if that name can't be inferred from the relation name. So has_many :photos will by default be linked to the Photo class, but if the real class name is e.g. CustomPhoto or namespaced Custom::Photo, you'll have to specify it with this option.
# app/models/custom/location.rb
module Custom
class Location < LHS::Record
endpoint '{+service}/locations'
endpoint '{+service}/locations/{id}'
has_many :photos, class_name: 'Custom::Photo'
end
end
# app/models/custom/photo.rb
module Custom
class Photo < LHS::Record
end
end
# app/models/transaction.rb
class Transaction < LHS::Record
endpoint '{+service}/transaction/{id}'
has_one :user
end
# app/models/user.rb
class User < LHS::Record
def email
self[:email_address]
end
end
# app/controllers/some_controller.rb
Transaction.find(1).user.email_address # steve@local.ch
GET https://service.example.com/transaction/1
{... user: { email_address: 'steve@local.ch' } }
class_name: Specify the class name of the relation. Use it only if that name can't be inferred from the relation name. So has_many :photos will by default be linked to the Photo class, but if the real class name is e.g. CustomPhoto or namespaced Custom::Photo, you'll have to specify it with this option.
# app/models/custom/location.rb
module Custom
class Location < LHS::Record
endpoint '{+service}/locations'
endpoint '{+service}/locations/{id}'
has_one :photo, class_name: 'Custom::Photo'
end
end
# app/models/custom/photo.rb
module Custom
class Photo < LHS::Record
end
end
If the actual item data is mixed with meta data in the response body, LHS allows you to configure a record in a way to automatically unwrap items from within nested response data.
item_key is used to unwrap the actual object from within the response body.
# app/models/location.rb
class Location < LHS::Record
configuration item_key: [:response, :location]
end
# app/controllers/some_controller.rb
location = Location.find(123)
location.id # 123
GET https://service.example.com/locations/123
{... response: { location: { id: 123 } } }
items_key key used to determine the collection of items of the current page (e.g. docs, items, etc.), defaults to 'items':
# app/models/search.rb
class Search < LHS::Record
configuration items_key: :docs
end
# app/controllers/some_controller.rb
search_result = Search.where(q: 'Starbucks')
search_result.first.address # Bahnhofstrasse 5, 8000 Zürich
GET https://service.example.com/search?q=Starbucks
{... docs: [... {... address: 'Bahnhofstrasse 5, 8000 Zürich' }] }
In order to load linked data from already retrieved data, you can use load! (or reload!).
# app/controllers/some_controller.rb
record = Record.find(1)
record.associated_thing.load!
GET https://things/4
{ name: "Steve" }
# app/controllers/some_controller.rb
record.associated_thing.name # Steve
record.associated_thing.load! # Does NOT create another request, as it is already loaded
record.associated_thing.reload! # Does request the data again from remote
GET https://things/4
{ name: "Steve" }
Method chaining, also known as named parameter idiom, is a common syntax for invoking multiple method calls in object-oriented programming languages. Each method returns an object, allowing the calls to be chained together without requiring variables to store the intermediate results
In order to simplify and enhance preparing complex queries for performing single or multiple requests, LHS implements query chains to find single or multiple records.
LHS query chains do lazy evaluation to only perform as many requests as needed, when the data to be retrieved is actually needed.
Any method, accessing the content of the data to be retrieved, is resolving the chain in place – like .each, .first, .some_attribute_name. Nevertheless, if you just want to resolve the chain in place, and nothing else, fetch should be the method of your choice:
# app/controllers/some_controller.rb
Record.where(color: 'blue').fetch
# app/controllers/some_controller.rb
records = Record.where(color: 'blue')
[...]
records.where(available: true).each do |record|
[...]
end
GET https://service.example.com/records?color=blue&available=true
In case you wan't to check/debug the current values for where in the chain, you can use where_values_hash:
records.where_values_hash
# {color: 'blue', available: true}
Some endpoints could respond only with a plain list of links and without any expanded data, like search results.
Use expanded to have LHS expand that data, by performing necessary requests in parallel:
# app/controllers/some_controller.rb
Search.where(what: 'Cafe').expanded
GET https://service.example.com/search?what=Cafe
{...
"items" : [
{"href": "https://service.example.com/records/1"},
{"href": "https://service.example.com/records/2"},
{"href": "https://service.example.com/records/3"}
]
}
In parallel:
> GET https://service.example.com/records/1
< {... name: 'Cafe Einstein'}
> GET https://service.example.com/records/2
< {... name: 'Starbucks'}
> GET https://service.example.com/records/3
< {... name: 'Plaza Cafe'}
{
...
"items" : [
{
"href": "https://service.example.com/records/1",
"name": 'Cafe Einstein',
...
},
{
"href": "https://service.example.com/records/2",
"name": 'Starbucks',
...
},
{
"href": "https://service.example.com/records/3",
"name": 'Plaza Cafe',
...
}
]
}
You can also apply request options to expanded. Those options will be used to perform the additional requests to expand the data:
# app/controllers/some_controller.rb
Search.where(what: 'Cafe').expanded(auth: { bearer: access_token })
One benefit of chains is lazy evaluation. But that also means they only get resolved when data is accessed. This makes it hard to catch errors with normal rescue blocks:
# app/controllers/some_controller.rb
def show
@records = Record.where(color: blue) # returns a chain, nothing is resolved, no http requests are performed
rescue => e
# never ending up here, because the http requests are actually performed in the view, when the query chain is resolved
end
# app/views/some/view.haml
= @records.each do |record| # .each resolves the query chain, leads to http requests beeing performed, which might raises an exception
= record.name
To simplify error handling with chains, you can also chain error handlers to be resolved, as part of the chain.
If you need to render some different view in Rails based on an LHS error raised during rendering the view, please proceed as following:
# app/controllers/some_controller.rb
def show
@records = Record
.handle(LHC::Error, ->(error){ handle_error(error) })
.where(color: 'blue')
render 'show'
render_error if @error
end
private
def handle_error(error)
@error = error
nil
end
def render_error
self.response_body = nil # required to not raise AbstractController::DoubleRenderError
render 'error'
end
> GET https://service.example.com/records?color=blue
< 406
In case no matching error handler is found the error gets re-raised.
-> Read more about LHC error types/classes
If you want to inject values for the failing records, that might not have been found, you can inject values for them with error handlers:
# app/controllers/some_controller.rb
data = Record
.handle(LHC::Unauthorized, ->(response) { Record.new(name: 'unknown') })
.find(1, 2, 3)
data[1].name # 'unknown'
In parallel:
> GET https://service.example.com/records/1
< 200
> GET https://service.example.com/records/2
< 400
> GET https://service.example.com/records/3
< 200
-> Read more about LHC error types/classes
If an error handler returns nil an empty LHS::Record is returned, not nil!
In case you want to ignore errors and continue working with nil in those cases,
please use ignore:
# app/controllers/some_controller.rb
record = Record.ignore(LHC::NotFound).find_by(color: 'blue')
record # nil
In case you need to resolve a query chain in place, use fetch:
# app/controllers/some_controller.rb
records = Record.where(color: 'blue').fetch
You can apply options to the request chain. Those options will be forwarded to the request perfomed by the chain/query:
# app/controllers/some_controller.rb
options = { auth: { bearer: '123456' } } # authenticated with OAuth token
# app/controllers/some_controller.rb
AuthenticatedRecord = Record.options(options)
# app/controllers/some_controller.rb
blue_records = AuthenticatedRecord.where(color: 'blue')
GET https://service.example.com/records?color=blue { headers: { 'Authentication': 'Bearer 123456' } }
# app/controllers/some_controller.rb
AuthenticatedRecord.create(color: 'red')
POST https://service.example.com/records { body: '{ color: "red" }' }, headers: { 'Authentication': 'Bearer 123456' } }
# app/controllers/some_controller.rb
record = AuthenticatedRecord.find(123)
GET https://service.example.com/records/123 { headers: { 'Authentication': 'Bearer 123456' } }
# app/controllers/some_controller.rb
authenticated_record = record.options(options) # starting a new chain based on the found record
# app/controllers/some_controller.rb
authenticated_record.valid?
POST https://service.example.com/records/validate { body: '{...}', headers: { 'Authentication': 'Bearer 123456' } }
# app/controllers/some_controller.rb
authenticated_record.save
POST https://service.example.com/records { body: '{...}', headers: { 'Authentication': 'Bearer 123456' } }
# app/controllers/some_controller.rb
authenticated_record.destroy
DELETE https://service.example.com/records/123 { headers: { 'Authentication': 'Bearer 123456' } }
# app/controllers/some_controller.rb
authenticated_record.update(name: 'Steve')
POST https://service.example.com/records/123 { body: '{...}', headers: { 'Authentication': 'Bearer 123456' } }
page sets the page that you want to request.
per sets the amount of items requested per page.
limit is an alias for per. But without providing arguments, it resolves the query and provides the current response limit per page
# app/controllers/some_controller.rb
Record.page(3).per(20).where(color: 'blue')
GET https://service.example.com/records?offset=40&limit=20&color=blue
# app/controllers/some_controller.rb
Record.page(3).per(20).where(color: 'blue')
GET https://service.example.com/records?offset=40&limit=20&color=blue
The applied pagination strategy depends on whats configured for the particular record: See Record pagination
You can configure pagination on a per record base. LHS differentiates between the pagination strategy (how items/pages are navigated and calculated) and pagination keys (how stuff is named and accessed).
The offset pagination strategy is LHS's default pagination strategy, so nothing needs to be (re-)configured.
The offset pagination strategy starts with 0 and offsets by the amount of items, thay you've already recived – typically limit.
# app/models/record.rb
class Search < LHS::Record
endpoint '{+service}/search'
end
# app/controllers/some_controller.rb
Record.all
GET https://service.example.com/records?limit=100
{
items: [{...}, ...],
total: 300,
limit: 100,
offset: 0
}
In parallel:
GET https://service.example.com/records?limit=100&offset=100
GET https://service.example.com/records?limit=100&offset=200
In comparison to the offset strategy, the page strategy just increases by 1 (page) and sends the next batch of items for the next page.
# app/models/record.rb
class Search < LHS::Record
configuration pagination_strategy: 'page', pagination_key: 'page'
endpoint '{+service}/search'
end
# app/controllers/some_controller.rb
Record.all
GET https://service.example.com/records?limit=100
{
items: [{...}, ...],
total: 300,
limit: 100,
page: 1
}
In parallel:
GET https://service.example.com/records?limit=100&page=2
GET https://service.example.com/records?limit=100&page=3
In comparison to the offset strategy, the start strategy indicates with which item the current page starts.
Typically it starts with 1 and if you get 100 items per page, the next start is 101.
# app/models/record.rb
class Search < LHS::Record
configuration pagination_strategy: 'start', pagination_key: 'startAt'
endpoint '{+service}/search'
end
# app/controllers/some_controller.rb
Record.all
GET https://service.example.com/records?limit=100
{
items: [{...}, ...],
total: 300,
limit: 100,
page: 1
}
In parallel:
GET https://service.example.com/records?limit=100&startAt=101
GET https://service.example.com/records?limit=100&startAt=201
limit_key sets the key used to indicate how many items you want to retrieve per page e.g. size, limit, etc.
In case the limit_key parameter differs for how it needs to be requested from how it's provided in the reponse, use body and parameter subkeys.
# app/models/record.rb
class Record < LHS::Record
configuration limit_key: { body: [:pagination, :max], parameter: :max }
endpoint '{+service}/records'
end
# app/controllers/some_controller.rb
records = Record.where(color: 'blue')
records.limit # 20
GET https://service.example.com/records?color=blue&max=100
{ ...
items: [...],
pagination: { max: 20 }
}
pagination_key defines which key to use to paginate a page (e.g. offset, page, startAt etc.).
In case the limit_key parameter differs for how it needs to be requested from how it's provided in the reponse, use body and parameter subkeys.
# app/models/record.rb
class Record < LHS::Record
configuration pagination_key: { body: [:pagination, :page], parameter: :page }, pagination_strategy: :page
endpoint '{+service}/records'
end
# app/controllers/some_controller.rb
records = Record.where(color: 'blue').all
records.length # 300
GET https://service.example.com/records?color=blue&limit=100
{... pagination: { page: 1 } }
In parallel:
GET https://service.example.com/records?color=blue&limit=100&page=2
{... pagination: { page: 2 } }
GET https://service.example.com/records?color=blue&limit=100&page=3
{... pagination: { page: 3 } }
total_key defines which key to user for pagination to describe the total amount of remote items (e.g. total, totalResults, etc.).
# app/models/record.rb
class Record < LHS::Record
configuration total_key: [:pagination, :total]
endpoint '{+service}/records'
end
# app/controllers/some_controller.rb
records = Record.where(color: 'blue').fetch
records.length # 100
records.count # 300
GET https://service.example.com/records?color=blue&limit=100
{... pagination: { total: 300 } }
next? Tells you if there is a next link or not.
# app/controllers/some_controller.rb
@records = Record.where(color: 'blue').fetch
GET https://service.example.com/records?color=blue&limit=100
{... items: [...], next: 'https://service.example.com/records?color=blue&limit=100&offset=100' }
# app/views/some_view.haml
- if @records.next?
= render partial: 'next_arrow'
previous? Tells you if there is a previous link or not.
# app/controllers/some_controller.rb
@records = Record.where(color: 'blue').fetch
GET https://service.example.com/records?color=blue&limit=100
{... items: [...], previous: 'https://service.example.com/records?color=blue&limit=100&offset=100' }
# app/views/some_view.haml
- if @records.previous?
= render partial: 'previous_arrow'
LHS implements an interface that makes it partially working with Kaminari.
The kaminari’s page parameter is in params[:page]. For example, you can use kaminari to render paginations based on LHS Records. Typically, your code will look like this:
# controller
@items = Record.page(params[:page]).per(100)
# view
= paginate @items
create will return the object in memory if persisting fails, providing validation errors in .errors (See record validation).
create! instead will raise an exception.
create always builds the data of the local object first, before it tries to sync with an endpoint. So even if persisting fails, the local object is build.
# app/controllers/some_controller.rb
record = Record.create(
text: 'Hello world'
)
POST https://service.example.com/records { body: "{ 'text' : 'Hello world' }" }
-> See record validation for how to handle validation errors when creating records.
item_created_key key used to merge record data thats nested in the creation response body:
# app/models/location.rb
class Location < LHS::Record
configuration item_created_key: [:response, :location]
end
# app/controllers/some_controller.rb
location.create(lat: '47.3920152', long: '8.5127981')
location.address # Förrlibuckstrasse 62, 8005 Zürich
POST https://service.example.com/locations { body: "{ 'lat': '47.3920152', long: '8.5127981' }" }
{... { response: { location: {... address: 'Förrlibuckstrasse 62, 8005 Zürich' } } } }
# app/models/restaurant.rb
class Restaurant < LHS::Record
endpoint '{+service}/restaurants/{id}'
end
# app/models/feedback.rb
class Feedback < LHS::Record
endpoint '{+service}/restaurants/{restaurant_id}/feedbacks'
end
# app/controllers/some_controller.rb
restaurant = Restaurant.find(1)
GET https://service.example.com/restaurants/1
{... reviews: { href: 'https://service.example.com/restaurants/1/reviews' }}
# app/controllers/some_controller.rb
restaurant.reviews.create(
text: 'Simply awesome!'
)
POST https://service.example.com/restaurants/1/reviews { body: "{ 'text': 'Simply awesome!' }" }
With new or build you can start building new records from scratch, which can be persisted with save:
# app/controllers/some_controller.rb
record = Record.new # or Record.build
record.name = 'Starbucks'
record.save
POST https://service.example.com/records { body: "{ 'name' : 'Starbucks' }" }
save persist the whole object in it's current state.
save will return false if persisting fails. save! instead will raise an exception.
# app/controllers/some_controller.rb
record = Record.find('1z-5r1fkaj')
GET https://service.example.com/records/1z-5r1fkaj
{ name: 'Starbucks', recommended: null }
# app/controllers/some_controller.rb
record.recommended = true
record.save
POST https://service.example.com/records/1z-5r1fkaj { body: "{ 'name': 'Starbucks', 'recommended': true }" }
-> See record validation for how to handle validation errors when updating records.
update persists the whole object after new parameters are applied through arguments.
update will return false if persisting fails. update! instead will raise an exception.
update always updates the data of the local object first, before it tries to sync with an endpoint. So even if persisting fails, the local object is updated.
# app/controllers/some_controller.rb
record = Record.find('1z-5r1fkaj')
GET https://service.example.com/records/1z-5r1fkaj
{ name: 'Starbucks', recommended: null }
# app/controllers/some_controller.rb
record.update(recommended: true)
POST https://service.example.com/records/1z-5r1fkaj { body: "{ 'name': 'Starbucks', 'recommended': true }" }
-> See record validation for how to handle validation errors when updating records.
partial_update updates just the provided parameters.
partial_update will return false if persisting fails. partial_update! instead will raise an exception.
partial_update always updates the data of the local object first, before it tries to sync with an endpoint. So even if persisting fails, the local object is updated.
# app/controllers/some_controller.rb
record = Record.find('1z-5r1fkaj')
GET https://service.example.com/records/1z-5r1fkaj
{ name: 'Starbucks', recommended: null }
# app/controllers/some_controller.rb
record.partial_update(recommended: true)
POST https://service.example.com/records/1z-5r1fkaj { body: "{ 'name': 'Starbucks', 'recommended': true }" }
-> See record validation for how to handle validation errors when updating records.
LHS injects parameters provided to create, update, partial_update, save etc. into an endpoint's URL when matching:
# app/models/feedback.rb
class Feedback << LHS::Record
endpoint '{+service}/records/{record_id}/feedbacks'
end
# app/controllers/some_controller.rb
Feedback.create(record_id: 51232, text: 'Great Restaurant!')
POST https://service.example.com/records/51232/feedbacks { body: "{ 'text' : 'Great Restaurant!' }" }
In order to validate records before persisting them, you can use the valid? (validate alias) method.
It's not recommended to validate records anywhere, including application side validation via ActiveModel::Validations, except, if you validate them via the same endpoint/service, that also creates them.
The specific endpoint has to support validations without persistence. An endpoint has to be enabled (opt-in) in your record configurations:
# app/models/user.rb
class User < LHS::Record
endpoint '{+service}/users', validates: { params: { persist: false } }
end
# app/controllers/some_controller.rb
user = User.build(email: 'i\'m not an email address')
unless user.valid?
@errors = user.errors
render 'new' and return
end
POST https://service.example.com/users?persist=false { body: '{ "email" : "i'm not an email address"}' }
{
"field_errors": [{
"path": ["email"],
"code": "WRONG_FORMAT",
"message": "The property value's format is incorrect."
}],
"message": "Email must have the correct format."
}
The functionalities of LHS::Errors pretty much follow those of ActiveModel::Validation:
# app/views/some_view.haml
@errors.any? # true
@errors.include?(:email) # true
@errors[:email] # ['WRONG_FORMAT']
@errors.messages # {:email=>["Translated error message that this value has the wrong format"]}
@errors.codes # {:email=>["WRONG_FORMAT"]}
@errors.message # Email must have the correct format."
The parameters passed to the validates endpoint option are used to perform record validations:
# app/models/user.rb
class User < LHS::Record
endpoint '{+service}/users', validates: { params: { persist: false } } # will add ?persist=false to the request
endpoint '{+service}/users', validates: { params: { publish: false } } # will add ?publish=false to the request
endpoint '{+service}/users', validates: { params: { validates: true } } # will add ?validates=true to the request
endpoint '{+service}/users', validates: { path: 'validate' } # will perform a validation via ...users/validate
end
The HTTP status code received from the endpoint when performing validations on a record, is available through the errors object:
# app/controllers/some_controller.rb
record.save
record.errors.status_code # 400
Clear the error messages like:
# app/controllers/some_controller.rb
record.errors.clear
In case you want to add application side validation errors, even though it's not recommended, do it as following:
user.errors.add(:name, 'WRONG_FORMAT')
If you work with complex data structures, you sometimes need to have validation errors delegated/scoped to nested data.
This features makes LHS::Records compatible with how Rails or Simpleform renders/builds forms and especially error messages:
# app/controllers/some_controller.rb
unless @customer.save
@errors = @customer.errors
end
POST https://service.example.com/customers { body: "{ 'address' : { 'street': 'invalid', housenumber: '' } }" }
{
"field_errors": [{
"path": ["address", "street"],
"code": "REQUIRED_PROPERTY_VALUE_INCORRECT",
"message": "The property value is incorrect."
},{
"path": ["address", "housenumber"],
"code": "REQUIRED_PROPERTY_VALUE",
"message": "The property value is required."
}],
"message": "Some data is invalid."
}
# app/views/some_view.haml
= form_for @customer, as: :customer do |customer_form|
= fields_for 'customer[:address]', @customer.address, do |address_form|
= fields_for 'customer[:address][:street]', @customer.address.street, do |street_form|
= street_form.input :name
= street_form.input :house_number
This would render nested forms and would also render nested form errors for nested data structures.
You can also access those nested errors like:
@customer.address.errors
@customer.address.street.errors
If a translation exists for one of the following translation keys, LHS will provide a translated error (also in the following order) rather than the plain error message/code, when building forms or accessing @errors.messages:
lhs.errors.records.<record_name>.attributes.<attribute_name>.<error_code>
e.g. lhs.errors.records.customer.attributes.name.unsupported_property_value
lhs.errors.records.<record_name>.<error_code>
e.g. lhs.errors.records.customer.unsupported_property_value
lhs.errors.messages.<error_code>
e.g. lhs.errors.messages.unsupported_property_value
lhs.errors.attributes.<attribute_name>.<error_code>
e.g. lhs.errors.attributes.name.unsupported_property_value
lhs.errors.fallback_message
lhs.errors.records.<record_name>.attributes.<collection>.<attribute_name>.<error_code>
e.g. lhs.errors.records.appointment_proposal.attributes.appointments.date_time.date_property_not_in_future
If an endpoint returns errors in the response body, that is enough to interpret it as: persistance failed. The response status code in this scenario is neglected.
In some cases, you need non blocking meta information about potential problems with the created record, so called warnings.
If the API endpoint implements warnings, returned when validating, they are provided just as errors (same interface and methods) through the warnings attribute:
# app/controllres/some_controller.rb
@presence = Presence.options(params: { synchronize: false }).create(
place: { href: 'http://storage/places/1' }
)
POST https://service.example.com/presences { body: '{ "place": { "href": "http://storage/places/1" } }' }
{
field_warnings: [{
code: 'WILL_BE_RESIZED',
path: ['place', 'photos', 0],
message: 'This photo is too small and will be resized.'
}
}
presence.warnings.any? # true
presence.place.photos[0].warnings.messages.first # 'This photo is too small and will be resized.'
ActiveModel::Validations none the lessIf you are using ActiveModel::Validations, even though it's not recommended, and you add errors to the LHS::Record instance, then those errors will be overwritten by the errors from ActiveModel::Validations when using save or valid?.
So in essence, mixing ActiveModel::Validations and LHS built-in validations (via endpoints), is not compatible, yet.
Rails form_for view-helper can be used in combination with instances of LHS::Records to autogenerate forms:
<%= form_for(@instance, url: '/create') do |f| %>
<%= f.text_field :name %>
<%= f.text_area :text %>
<%= f.submit "Create" %>
<% end %>
destroy deletes a record.
# app/controllers/some_controller.rb
record = Record.find('1z-5r1fkaj')
GET https://service.example.com/records/1z-5r1fkaj
# app/controllers/some_controller.rb
record.destroy
DELETE https://service.example.com/records/1z-5r1fkaj
You can also destroy records directly without fetching them first:
# app/controllers/some_controller.rb
destroyed_record = Record.destroy('1z-5r1fkaj')
DELETE https://service.example.com/records/1z-5r1fkaj
or with parameters:
# app/controllers/some_controller.rb
destroyed_records = Record.destroy(name: 'Steve')
DELETE https://service.example.com/records?name='Steve'
Sometimes it is neccessary to implement custom getters and setters and convert data to a processable (endpoint) format behind the scenes.
You can define setter methods in LHS::Records that will be used by initializers (new) and setter methods, that convert data provided, before storing it in the record and persisting it with a remote endpoint:
# app/models/user.rb
class Feedback < LHS::Record
def ratings=(values)
super(
values.map { |k, v| { name: k, value: v } }
)
end
end
# app/controllers/some_controller.rb
record = Record.new(ratings: { quality: 3 })
record.ratings # [{ :name=>:quality, :value=>3 }]
If you implement accompanying getter methods, the whole data conversion would be internal only:
# app/models/user.rb
class Feedback < LHS::Record
def ratings=(values)
super(
values.map { |k, v| { name: k, value: v } }
)
end
def ratings
super.map { |r| [r[:name], r[:value]] }]
end
end
# app/controllers/some_controller.rb
record = Record.new(ratings: { quality: 3 })
record.ratings # {:quality=>3}
In a service-oriented architecture using hyperlinks/hypermedia, records/resources can contain hyperlinks to other records/resources.
When fetching records with LHS, you can specify in advance all the linked resources that you want to include in the results.
With includes or includes_all (to enforce fetching all remote objects for paginated endpoints), LHS ensures that all matching and explicitly linked resources are loaded and merged.
Including linked resources/records is heavily influenced by https://guides.rubyonrails.org/active_record_querying.html and you should read it to understand this feature in all it's glory.
In case endpoints are paginated and you are certain that you'll need all objects of a set and not only the first page/batch, use includes_all.
LHS will ensure that all linked resources are around by loading all pages (parallelized/performance optimized).
# app/controllers/some_controller.rb
customer = Customer.includes_all(contracts: :products).find(1)
> GET https://service.example.com/customers/1
< {... contracts: { href: 'https://service.example.com/customers/1/contracts' } }
> GET https://service.example.com/customers/1/contracts?limit=100
< {... items: [...], limit: 10, offset: 0, total: 32 }
In parallel:
> GET https://service.example.com/customers/1/contracts?limit=10&offset=10
< {... products: [{ href: 'https://service.example.com/product/LBC' }] }
> GET https://service.example.com/customers/1/contracts?limit=10&offset=20
< {... products: [{ href: 'https://service.example.com/product/LBB' }] }
In parallel:
> GET https://service.example.com/product/LBC
< {... name: 'Local Business Card' }
> GET https://service.example.com/product/LBB
< {... name: 'Local Business Basic' }
# app/controllers/some_controller.rb
customer.contracts.length # 32
customer.contracts.first.products.first.name # Local Business Card
includes includes the first page/response when loading the linked resource. If the endpoint is paginated, only the first page will be included.
# app/controllers/some_controller.rb
customer = Customer.includes(contracts: :products).find(1)
> GET https://service.example.com/customers/1
< {... contracts: { href: 'https://service.example.com/customers/1/contracts' } }
> GET https://service.example.com/customers/1/contracts?limit=100
< {... items: [...], limit: 10, offset: 0, total: 32 }
In parallel:
> GET https://service.example.com/product/LBC
< {... name: 'Local Business Card' }
> GET https://service.example.com/product/LBB
< {... name: 'Local Business Basic' }
# app/controllers/some_controller.rb
customer.contracts.length # 10
customer.contracts.first.products.first.name # Local Business Card
The method syntax of includes and includes_all, allows you include hyperlinks stored in deep nested data strutures:
Some examples:
Record.includes(:localch_account, :entry)
# Includes localch_account -> entry
# { localch_account: { href: '...', entry: { href: '...' } } }
Record.includes([:localch_account, :entry])
# Includes localch_account and entry
# { localch_account: { href: '...' }, entry: { href: '...' } }
Record.includes(campaign: [:entry, :user])
# Includes campaign and entry and user from campaign
# { campaign: { href: '...' , entry: { href: '...' }, user: { href: '...' } } }
When including linked resources with includes or includes_all, already defined records and their endpoints and configurations are used to make the requests to fetch the additional data.
That also means that options for endpoints of linked resources are applied when requesting those in addition.
This applies for example a records endpoint configuration even though it's fetched/included through another record:
# app/models/favorite.rb
class Favorite < LHS::Record
endpoint '{+service}/users/{user_id}/favorites', auth: { basic: { username: 'steve', password: 'can' } }
endpoint '{+service}/users/{user_id}/favorites/:id', auth: { basic: { username: 'steve', password: 'can' } }
end
# app/models/place.rb
class Place < LHS::Record
endpoint '{+service}/v2/places', auth: { basic: { username: 'steve', password: 'can' } }
endpoint '{+service}/v2/places/{id}', auth: { basic: { username: 'steve', password: 'can' } }
end
# app/controllers/some_controller.rb
Favorite.includes(:place).where(user_id: current_user.id)
> GET https://service.example.com/users/123/favorites { headers: { 'Authentication': 'Basic c3RldmU6Y2Fu' } }
< {... items: [... { place: { href: 'https://service.example.com/place/456' } } ] }
In parallel:
> GET https://service.example.com/place/456 { headers: { 'Authentication': 'Basic c3RldmU6Y2Fu' } }
> GET https://service.example.com/place/789 { headers: { 'Authentication': 'Basic c3RldmU6Y2Fu' } }
> GET https://service.example.com/place/1112 { headers: { 'Authentication': 'Basic c3RldmU6Y2Fu' } }
> GET https://service.example.com/place/5423 { headers: { 'Authentication': 'Basic c3RldmU6Y2Fu' } }
Use references to apply request options to requests performed to fetch included records:
# app/controllers/some_controller.rb
Favorite.includes(:place).references(place: { auth: { bearer: '123' }}).where(user_id: 1)
GET https://service.example.com/users/1/favorites
{... items: [... { place: { href: 'https://service.example.com/places/2' } }] }
In parallel:
GET https://service.example.com/places/2 { headers: { 'Authentication': 'Bearer 123' } }
GET https://service.example.com/places/3 { headers: { 'Authentication': 'Bearer 123' } }
GET https://service.example.com/places/4 { headers: { 'Authentication': 'Bearer 123' } }
Be careful using methods for batch processing. They could result in a lot of HTTP requests!
all fetches all records from the service by doing multiple requests, best-effort parallelization, and resolving endpoint pagination if necessary:
records = Record.all
> GET https://service.example.com/records?limit=100
< {...
items: [...]
total: 900,
limit: 100,
offset: 0
}
In parallel:
> GET https://service.example.com/records?limit=100&offset=100
> GET https://service.example.com/records?limit=100&offset=200
> GET https://service.example.com/records?limit=100&offset=300
> GET https://service.example.com/records?limit=100&offset=400
> GET https://service.example.com/records?limit=100&offset=500
> GET https://service.example.com/records?limit=100&offset=600
> GET https://service.example.com/records?limit=100&offset=700
> GET https://service.example.com/records?limit=100&offset=800
all is chainable and has the same interface like where:
Record.where(color: 'blue').all
Record.all.where(color: 'blue')
Record.all(color: 'blue')
All three are doing the same thing: fetching all records with the color 'blue' from the endpoint while resolving pagingation if endpoint is paginated.
In case an API does not provide pagination information in the repsponse data (limit, offset and total), LHS keeps on loading pages when requesting all until the first empty page responds.
find_each is a more fine grained way to process single records that are fetched in batches.
Record.find_each(start: 50, batch_size: 20, params: { has_reviews: true }, headers: { 'Authorization': 'Bearer 123' }) do |record|
# Iterates over each record. Starts with record no. 50 and fetches 20 records each batch.
record
break if record.some_attribute == some_value
end
find_in_batches is used by find_each and processes batches.
Record.find_in_batches(start: 50, batch_size: 20, params: { has_reviews: true }, headers: { 'Authorization': 'Bearer 123' }) do |records|
# Iterates over multiple records (batch size is 20). Starts with record no. 50 and fetches 20 records each batch.
records
break if records.first.name == some_value
end
Based on ActiveRecord's implementation, LHS implements becomes, too.
It's a way to convert records of a certain type A to another certain type B.
NOTE: RPC-style actions, that are discouraged in REST anyway, are utilizable with this functionality, too. See the following example:
# app/models/location.rb
class Location < LHS::Record
endpoint '{+service}/locations'
endpoint '{+service}/locations/{id}'
end
# app/models/synchronization.rb
class Synchronization < LHS::Record
endpoint '{+service}/locations/{id}/sync'
end
# app/controllers/some_controller.rb
location = Location.find(1)
GET https://service.example.com/location/1
# app/controllers/some_controller.rb
synchronization = location.becomes(Synchronization)
synchronization.save!
POST https://service.example.com/location/1/sync { body: '{ ... }' }
Allows you to set the attributes by passing in a hash of attributes.
entry = LocalEntry.new
entry.assign_attributes(company_name: 'localsearch')
entry.company_name # => 'localsearch'
By default, LHS does not perform the same http request multiple times during one request/response cycle.
# app/models/user.rb
class User < LHS::Record
endpoint '{+service}/users/{id}'
end
# app/models/location.rb
class Location < LHS::Record
endpoint '{+service}/locations/{id}'
end
# app/controllers/some_controller.rb
def index
@user = User.find(1)
@locations = Location.includes(:owner).find(2)
end
GET https://service.example.com/users/1
GET https://service.example.com/location/2
{... owner: { href: 'https://service.example.com/users/1' } }
From cache:
GET https://service.example.com/users/1
It uses the LHC Caching Interceptor as caching mechanism base and sets a unique request id for every request cycle with Railties to ensure data is just cached within one request cycle and not shared with other requests.
Only GET requests are considered for caching by using LHC Caching Interceptor's cache_methods option internally and considers request headers when caching requests, so requests with different headers are not served from cache.
The LHS Request Cycle Cache is opt-out, so it's enabled by default and will require you to enable the LHC Caching Interceptor in your project.
By default the LHS Request Cycle Cache will use ActiveSupport::Cache::MemoryStore as its cache store. Feel free to configure a cache that is better suited for your needs by:
# config/initializers/lhs.rb
LHS.configure do |config|
config.request_cycle_cache = ActiveSupport::Cache::MemoryStore.new
end
If you want to disable the LHS Request Cycle Cache, simply disable it within configuration:
# config/initializers/lhs.rb
LHS.configure do |config|
config.request_cycle_cache_enabled = false
end
LHS supports tracing the source (in your application code) of http requests being made with methods like find find_by find_by! first first! last last!.
Following links, and using includes are not traced (just yet).
In order to enable tracing you need to enable it via LHS configuration:
# config/initializers/lhs.rb
LHS.configure do |config|
config.trace = Rails.env.development? || Rails.logger.level == 0 # debug
end
# app/controllers/application_controller.rb
code = Code.find(code: params[:code])
Called from onboarding/app/controllers/concerns/access_code_concern.rb:11:in `access_code'
However, following links and includes won't get traced (just yet):
# app/controllers/application_controller.rb
code = Code.includes(:places).find(123)
# Nothing is traced
{
places: [...]
}
code.places
{
token: "XYZABCDEF",
places:
[
{ href: "http://storage-stg.preprod-local.ch/v2/places/egZelgYhdlg" }
]
}
Best practice in regards of testing applications using LHS, is to let LHS fetch your records, actually perform HTTP requests and WebMock to stub/mock those http requests/responses.
This follows the Black Box Testing approach and prevents you from creating constraints to LHS' internal structures and mechanisms, which will break as soon as we change internals.
# specs/*/some_spec.rb
let(:contracts) do
[
{number: '1'},
{number: '2'},
{number: '3'}
]
end
before do
stub_request(:get, "https://service.example.com/contracts")
.to_return(
body: {
items: contracts,
limit: 10,
total: contracts.length,
offset: 0
}.to_json
)
end
it 'displays contracts' do
visit 'contracts'
contracts.each do |contract|
expect(page).to have_content(contract[:number])
end
end
In order to not run into caching issues during your tests, when (request cycle cache)[#request-cycle-cache] is enabled, simply require the following helper in your tests:
# spec/spec_helper.rb
require 'lhs/test/request_cycle_cache_helper'
This will initialize a MemoryStore cache for LHC::Caching interceptor and resets the cache before every test.
Use fetch in tests to resolve chains in place and expect WebMock stubs to be requested.
# specs/*/some_spec.rb
records = Record.where(color: 'blue').where(available: true).where(color: 'red')
expect(
records.fetch
).to have_requested(:get, %r{records/})
.with(query: hash_including(color: 'blue', available: true))
As where chains are not resolving to HTTP-requests when no data is accessed, you can use where_values_hash to access the values that would be used to resolve the chain, and test those:
# specs/*/some_spec.rb
records = Record.where(color: 'blue').where(available: true).where(color: 'red')
expect(
records.where_values_hash
).to eq {color: 'red', available: true}
FAQs
Unknown package
We found that lhs demonstrated a not healthy version release cadence and project activity because the last version was released a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
In this segment of the Risky Business podcast, Feross Aboukhadijeh and Patrick Gray discuss the challenges of tracking malware discovered in open source softare.
Research
Security News
A threat actor's playbook for exploiting the npm ecosystem was exposed on the dark web, detailing how to build a blockchain-powered botnet.
Security News
NVD’s backlog surpasses 20,000 CVEs as analysis slows and NIST announces new system updates to address ongoing delays.